QUANTITATIVE VALIDATION OF OPTICAL FLOW BASED MYOCARDIAL STRAIN MEASURES USING SONOMICROMETRY

Area of Science:

• Cardiovascular Imaging
• Biomedical Engineering
• Medical Ultrasound

Background:

• Current 2D strain measures inadequately represent complex 3D cardiac motion.
• 4D ultrasound technology captures intricate cardiac dynamics in a single dataset.
• Accurate quantification of myocardial strain is crucial for assessing cardiac function.

Purpose of the Study:

• To validate a 4D optical flow-based motion tracking algorithm for dynamic cardiac strain estimation.
• To compare the accuracy of the 4D ultrasound tracking method against sonomicrometry.
• To establish the in-vivo performance of a novel cardiac motion analysis technique.

Main Methods:

• Development of a 4D optical flow algorithm for motion tracking in 4D ultrasound data.
• In-vivo experiments on canine hearts involving coronary artery occlusion.
• Acquisition of simultaneous 4D ultrasound and sonomicrometry data.
• Direct comparison of estimated principal strain measures from the algorithm with sonomicrometry recordings.

Main Results:

• The 4D optical flow tracking method demonstrated strong agreement with sonomicrometry data.
• Accurate estimation of dynamic cardiac strain metrics was achieved using the 4D ultrasound approach.
• The study successfully validated the algorithm's performance in a relevant in-vivo setting.

Conclusions:

• The 4D optical flow-based strain estimation for 4D ultrasound is a validated and accurate method.
• This represents the first in-vivo validation study comparing optical flow strain estimation in 4D ultrasound with sonomicrometry.
• The findings support the clinical potential of advanced 4D ultrasound techniques for comprehensive cardiac function assessment.